SCOT: Multi-Source Cross-City Transfer with Optimal-Transport Soft-Correspondence Objective
This addresses the challenge of label scarcity in urban prediction tasks by enabling more stable and interpretable multi-source transfer, though it is incremental in refining existing methods.
The paper tackles the problem of cross-city transfer learning when cities have incompatible region partitions and no ground-truth correspondences, proposing SCOT which learns explicit soft correspondences via optimal transport and improves transfer accuracy and robustness across real-world tasks.
Cross-city transfer improves prediction in label-scarce cities by leveraging labeled data from other cities, but it becomes challenging when cities adopt incompatible partitions and no ground-truth region correspondences exist. Existing approaches either rely on heuristic region matching, which is often sensitive to anchor choices, or perform distribution-level alignment that leaves correspondences implicit and can be unstable under strong heterogeneity. We propose SCOT, a cross-city representation learning framework that learns explicit soft correspondences between unequal region sets via Sinkhorn-based entropic optimal transport. SCOT further sharpens transferable structure with an OT-weighted contrastive objective and stabilizes optimization through a cycle-style reconstruction regularizer. For multi-source transfer, SCOT aligns each source and the target to a shared prototype hub using balanced entropic transport guided by a target-induced prototype prior. Across real-world cities and tasks, SCOT consistently improves transfer accuracy and robustness, while the learned transport couplings and hub assignments provide interpretable diagnostics of alignment quality.